Diagonal Pixel Binning for Image Sensor Color Alignment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Pixel binning in image sensors often results in misalignment of color clusters, leading to spatial color artifacts and reduced image quality, especially when summing pixels orthogonally, which degrades the signal-to-noise ratio and increases readout noise.
Innovation Solution
Implementing a method that sums charge diagonally in image sensors, allowing for digital combination of signals from diagonally adjacent pixels, and enabling orthogonal binning by summing charge in respective sense regions and combining voltage signals on output lines, thereby improving image quality and spatial resolution.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If pixels are summed orthogonally in the pixel array, then readout time is reduced, but spatial color artifacts occur due to misalignment of color clusters
Solution Approach 1:
The patent transitions from orthogonal (horizontal/vertical) binning to diagonal binning, changing the dimension of summation from axis-aligned to diagonal directions. This dimensional change allows color clusters to be properly aligned while still reducing the number of readout operations, thereby maintaining both speed and spatial accuracy.
Solution Approach 2:
The patent introduces asymmetric binning patterns where different color planes (R, G, B) are binned along different diagonal directions rather than using uniform orthogonal binning. This asymmetric approach compensates for the different spatial relationships between color planes in the Bayer pattern, preventing color misalignment artifacts.
2Reliability
If pixels are summed after readout, then all pixels can be read out, but noise from readout circuitry is summed with pixel signals decreasing signal to noise ratio
Solution Approach 1:
The patent performs charge summation in the pixel array before the readout operation, rather than summing after readout. This preliminary action ensures that only the necessary binned signals are read out, eliminating the accumulation of readout circuitry noise while maintaining complete signal representation.
3Productivity
If orthogonal binning is used, then fewer pixels are read out, but clusters are out of alignment producing spatial color artifacts
Solution Approach 1:
The patent changes the binning dimension from orthogonal (horizontal/vertical) to diagonal directions. This allows color clusters to be properly aligned according to the Bayer pattern geometry while still reducing the number of readout operations, thereby maintaining both readout efficiency and color alignment precision.
Data Source
AI summary
Pixel binning is performed by summing charge from some pixels positioned diagonally in a pixel array. Pixel signals output from pixels positioned diagonally in the pixel array may be combined on the output lines. A signal representing summed charge produces a binned 2×1 cluster. A signal representing combined voltage signals produces a binned 2×1 cluster. A signal representing summed charge and a signal representing combined pixel signals can be combined digitally to produce a binned 2×2 pixel. Orthogonal binning may be performed on other pixels in the pixel array by summing charge on respective common sense regions and then combining the voltage signals that represent the summed charge on respective output lines.